Male circumcision (MC) is currently one of the most effective biomedical HIV-prevention strategies that can be implemented in high HIV prevalence countries (Auvert, et al, 2005; Gray, et al., 2007; Bailey, et al., 2007). Modeling shows that high uptake of MC must occur in a young adult male population to have the largest impact on the HIV/AIDS epidemic, but this may be offset by risk compensation (RC) (Bollinger et al., 2009; Gray, et al., 2007; UNAIDS, 2009); that is, men who are circumcised may compensate for the protection MC offers by increasing their risk behavior (Templeton, 2010). Follow-up with the three MC randomized control trial (RCT) cohorts in Kenya, Uganda, and South Africa suggests little RC over time. However, the men in these RCTs were followed intensely (every 3 or 6 months) with prevention counseling, STD treatment, and free condoms offered at each follow-up visit. Thus, the RCT cohorts are unlikely to provide a good indication of the degree to which RC will occur in an ordinary roll out of MC where repeat intensive prevention counseling is not feasible. There is a clear need to determine whether risk compensation occurs among men getting circumcised through national MC programs that are being implemented in high HIV-prevalence countries, and to develop strategies to prevent it. Since Zimbabwe is currently implementing an MC program as part of its national HIV prevention strategy, this is the ideal time and setting to conduct a longitudinal study to determine: a) whether RC occurs among men circumcised through this program, b) the prevalence and change in prevalence in RC over time since circumcision, and c) the behavioral determinants of RC. Thus, we propose to conduct a cohort study where we will accrue 1200 randomly selected men from those who have made a decision to get circumcised and a non-equivalent comparison group of 1200 men who have decided not to get circumcised. We will conduct surveys of these men at baseline (prior to circumcision) and at 6-, 12-, 24-, and 36-months, in order to determine whether RC occurs soon after MC healing or whether RC continues longer term. The Integrated Behavioral Model (Montano & Kasprzyk, 2008) will provide the theoretical framework for designing the survey to measure potential environmental (socio-cultural) and individual (attitudinal, normative, personal agency, sexual satisfaction) determinants of RC behavior. Analyses will be carried out to determine whether there is differential change in risk behavior among men who get circumcised compared with those who do not, and the patterns of change in RC over time; and to identify specific model constructs and beliefs that best explain RC. These results will then govern theory-based prevention messages. The study will include men in Zimbabwe's two main cities, Harare and Bulawayo, which will allow us to determine whether there are ethnic (Shona, Matabele) differences in RC and determinants. This study will be the first to provide information on the prevalence of risk compensation in a national program rolling out circumcision, as well as model the determinants of this behavior in order to design effective risk-prevention messages.